[SynSem] Merve Yazar / Daniel Saeger

retour Séminaire Syntaxe et Sémantique

Merve Yazar (U. Paris 8 & U. Chicago) - Sluicing Constructions in Romeyka

Daniel Saeger (U. Paris 8, SFL) Unaccusativity and light verb syntax: the case of ficar in Brazilian Portuguese

abstracts below

Merve Yazar

Sluicing Constructions in Romeyka

First observed by Ross (1969), sluicing is a type of ellipsis where an interrogative embedded clause is reduced to an overt wh-phrase as in (1):

1. Somebody just left- guess who [just left]. (Ross, 1969: 252)

It is assumed that derivation of such structures involves the wh-element moving to Spec CP, with the embedded TP being elided for wh-fronting languages such as English (Ross 1969, Merchant 2001). However, sluicing constructions in wh-in-situ languages pose challenges to such movement-deletion approaches.

Romeyka, an Asia Minor Greek variety, has been in contact with Turkish, a wh-in-situ language (Akar, 1999). Although the language is reported to be wh-fronting (Michelioudakis & Sitaridou, 2016), with an underlying VO order (Neocleous, 2020), I show that heritage speakers of Romeyka (n=3), being Turkish-dominant, exhibit patterns influenced by Turkish's wh-in-situ syntax. This influence is reflected in their tendency to prefer OV order as a default, parallel to Turkish. Regular wh-questions in these languages are exemplified in (2-3). Notably, given the unmarked SOV order, wh-elements in this variety don’t undergo obligatory movement and are subject to discourse-related scrambling, a pattern shared with Turkish (Akar, 1999).

2. Esi       tina                  idhes?             
   You     who.ACC        see.PST.2.SG
   ‘Who did you see?’                                                                             (Romeyka)
3. Sen      kim-i               gör-dü-n?
   You     who-ACC        see-PST-2.SG
   ‘Who did you see?’                                                                             (Turkish)

Heritage Romeyka also allows sluicing constructions, exemplified in (4), which raise questions about the derivation of such structures, given the wh-patterns in the variety.

4. I                                   Ayshe                          katina                          idhen              
   the.F.NOM.SG           Ayshe.F.NOM.SG      someone.ACC             see.PST.3.SG
   ama     tina                  uch                              eksero.
   but       who.ACC        NEG                            know.PRES.1.SG
   ‘Ayshe saw someone but I don’t know who.’

In this talk, I present a preliminary analysis of sluicing constructions as in (4) in heritage Romeyka. I show that the variety exhibits the three types of sluicing identified by Chung, Ladusaw, & McCloskey (1995). Based on preliminary data revealing connectivity effects between the wh-remnant and its correlate in the antecedent clause, such as case matching and parallelism in preposition stranding in both sluicing and non-sluicing environments, I argue that sluicing in this variety involves movement of the wh-remnant. Given the focus-bearing nature of wh-remnants in sluicing constructions, I suggest the movement is motivated by focus, parallel to Ince (2009)’s analysis of Turkish sluicing. Following Lobeck (1995) and Ince (2009), I assume the wh-element moves to FocP to satisfy EPP on Foc Head, thus bearing focus here. I further assume Foc Head carries Merchant (2008)’s ellipsis feature [E], triggering the deletion of the TP complement. This movement differs from regular wh-movement in that the ellipsis-licensing head carries an EPP feature, making the movement obligatory. Non-movement approaches to sluicing, such as those by Chung, Ladusaw, & McCloskey (1995), fail to account for these observations, which I will explore in detail. I conclude by discussing ongoing work on island effects in sluicing, which may further challenge and refine this analysis.

Daniel Saeger

Unaccusativity and light verb syntax: the case of ficar in Brazilian Portuguese

This talk examines the syntax of the verb ficar in Brazilian Portuguese (BP), focusing on its grammaticalization process and subsequent functional expansion into an existential verb, auxiliary verb, copula, and control verb. Through this evolution, ficar has developed the capacity to convey meanings ranging from inchoative and iterative aspects to irrealis mood. My investigation therefore centers on the polysemy of ficar, exploring its grammaticalization path and the mechanisms enabling its semantic versatility, with particular emphasis on identifying the underlying commonalities across its various meanings.
    Drawing on Marantz’s (1984) event decomposition framework, I propose that ficar is fundamentally an autocausative verb that evolved from a causative use (a putting verb, cf. Levin et al., 1995). I argue that most constructions involving ficar share two core syntactic components—VoiceP and vP—as integral elements of its internal structure. The diverse semantic interpretations appear to emerge from the interaction between these projections and their complements, with variations particularly evident when the complement is a verb phrase, where meaning differences correlate with the matrix verb’s event structure.
    The analysis extends to a comparative examination of ficar and another significant BP light verb, estar, revealing both shared properties and crucial distinctions. Key areas of discussion include the syntactic encoding of change-of-state predicates (Alexiadou et al., 2006; Schäfer, 2008); the (in)compatibility of stage-level and individual-level predicates (Kratzer, 1995) with these two verbs; the realization of iterative, durative, and inchoative aspects through gerunds; and the relationship between embedded CPs and irrealis mood expression (Stowell, 1982). The talk concludes with a cross-linguistic investigation of similar phenomena, reflecting my current research, aimed at identifying common patterns and deriving broader linguistic generalizations.